Mark Rober just set up one of the most interesting self-driving tests of 2025, and he did it by imitating Looney Tunes. The former NASA engineer and current YouTube mad scientist recreated the classic gag where Wile E. Coyote paints a tunnel onto a wall to fool the Road Runner.
Only this time, the test subject wasn’t a cartoon bird… it was a self-driving Tesla Model Y.
The result? A full-speed, 40 MPH impact straight into the wall. Watch the video and tell us what you think!
Don’t want to rock the boat but apart from being a you tube money earner this doesn’t prove or disprove anything. A lot of humans would be fooled by this also.
I am suspicious of the way the polystyrene wall broke in cartoon like shagged edges, almost like they were precut.
The point of the test is to demonstrate that vision-only, which Tesla has adopted is inadequate. A car with lidar or radar would have been able to “see” that the car was approaching an obstacle without being fooled by the imagary.
So yes, it seems a bit silly, but the underlying point is legitimate. If the software is fooled by this, then can you ever fully trust it? Especially when sensor systems exist that don’t have this problem at all. Would you want to be a pedestrian in a crosswalk with this car bearing down on you in FSD?
What would the wall being precut have to do with the car deciding to drive through it?
Nothing much is real anymore on YT
They were expecting this result to be possible. What were they supposed to do? Slam the car into the side if a building?
Yes but the main point that has been shown is that putting a screen up with the exact copy of the road and surroundings behind the screen is a daft and dangerous idea. It would be a better test if they had put up a polystyrene tree in the middle of the road and then checked if the car stopped.
I have never driven through a polystyrene wall with a picture of a road on it in 40 years because people just don’t put those things up, they don’t grow on roads etc etc.
Great YT clip for entertainment though.
The other two tests that the Tesla failed were more realistic. Heavy fog and heavy rain.
It failed both.
If your self driving car cannot handle weather, then it’s not self driving at all.
Actual lidar isn’t fooled by weather. Shitty optical only cameras are.
Agreed and that’s a real world scenario that is being tested which has real value.
Have you never heard of a mural before?
I have never seen a mural on a road depicting a road that is identical to the road that I am driving on. Hope that helps.
It may not rise to the level of proof, but it is a memorable and easily understood demonstration of something already proven by car safety researchers, as mentioned in the article.
Why shouldn’t they precut the wall into cartoony shapes? It adds entertainment and doesn’t compromise the demonstration.
Yep agreed. Having used Teslas adaptive cruise control I wouldn’t ever use self driving, not that I have it, unless I had a death wish. Quite honestly my previous Chinese MG was a lot less likely to kill me.